Page 102 - Tunable Lasers Handbook

4 CO, Isotope Lasers and Their Applications 83

(v = v,), a narrowband resonant dip appears in the intensity of the laser output
power, because the traveling-wave components constituting the standing-wave
field interact with the same group of molecules (or atoms). namely, those that
have zero velocity in the direction of the laser's optical axis (k.is=O). This dip in
the laser's output power was first verified by Szoke and Javan in the output of a
He-Ne laser at 1.15 pm [66]. and was appropriately named "Lamb-dip" since
Lamb predicted its occurrence.
An even more useful variant of the standing-wave saturation resonance was
first demonstrated by Lee and Skolnick [67] who inserted a low-pressure
absorber gas cell, which had a resonantly interacting absorption line, within the
standing-wave field of the laser's optical cavity. In this case the narrow resonant
change appeared as a "pip" increase in the laser's output power and was named
"inverted Lamb-dip."
To a very good first approximation. the line shapes and FWHM widths of
the Lamb-dips and the inverted Lamb-dips are determined by collision broaden-
ing and thus have Lorentzian profiles. In actual practice the absorber gas refer-
ence cells can be effectively used with much lower pressure gas fills than the
typical mixture pressures required to operate gas lasers. Thus both in principle
and in practice the long-term frequency stabilization techniques utilizing the
inverted Lamb-dip can provide much better frequency discriminators than those
using the Lamb-dip.
One of the best known early examples of inverted Lamb-dip stabilization is the
methane-(CH,) stabilized He-Ne laser oscillating at 3.39 pm. This absorber-laser
combination was first suggested and demonstrated by Shimoda in 1968 [68] and
was also extensively studied and utilized by Barger and Hall [69].
In the case of the CO, laser system the initial attempts to use CO, itself as a
reference via either the Lamb-dip or the inverted Lamb-dip techniques were not
very successful. Lamb-dip was only obtained with very low-pressure laser gas
fills and was prone to severe asymmetrical distortions due to competition from
adjacent transitions [70]. The inverted Lamb-dip stabilization method on the other
hand required very long (-1.7-m) CO, absorption cells heated to several hundred
(400°C) degrees above room temperature [71]. The poor results were due to the
fact that the lower state rotational-vibrational levels of the CO, laser transitions
do not belong to the ground state. and therefore the absorption coefficient of low-
pressure room-temperature CO, at 10 ym is very small. The small absorption
coefficient in turn made it difficult to observe and utilize directly the inverted
Lamb-dip resonance in the full-power output of the CO, laser. These difficulties
were overcome at Lincoln Laboratory in 1970, when. atthe suggestion of Javan.
we (Freed and Javan) first demonstrated [18] that excellent long-term frequenc?
stability and reproducibility of CO, lasers can be readily obtained (and greatly
improved on if necessary) by the frequency stabilization of the lasers to the srand-
ing-wave saturation resonance observed in the 1.3-ym upper-state-to-ground-state
fluorescence of CO,. as graphically illustrated in Fig. 8.